This invention relates to an apparatus and method for detecting flaws in the surface and sub-surface region of a material by inducing an eddy current therein at a relatively high frequency.
The ability of an eddy current probe to detect flaws in the surface and sub-surface region of materials depends upon the resistivity of the material and the excitation frequency applied to the coil within the eddy current probe. The resistivity of a material provides a classification of the material as conducting or semi-conducting, hence, a material having a relatively high value of resistivity is commonly recognized as semi-conducting in nature. To nondestructively test a semi-conducting material, a relatively high frequency is necessary to induce an eddy current within the material. For example, to detect a hair line surface crack in a block of carbon, the eddy current probe must have its coil excited at a relatively high frequency, a frequency which approaches 6 megahertz.
The excitation frequency of the coil also affects the depth of penetration of the eddy currents in the material under test. As commonly recognized in the nondestructive testing art, the higher the frequency of excitation, the lesser the depth of penetration within the sub-surface region of the material under test. If the depth of penetration in the above example is limited to approximately 35 mils, the coil must be excited at a frequency equal to or greater than 7 megahertz.
High sensitivity, commercially available, eddy current probes have an upper frequency range limitation of 6 megahertz. An example of one high frequency eddy current probe is the SPO2000 Probe manufactured by Nortec Inc. of Kennewick, Wash. These high sensitivity probes generally have fine wire wound about a small bobbin. To achieve the high sensitivity to detect the hair line surface crack in the carbon block example, the coil therein must be excited by a signal approaching or exceeding 7 megahertz. At that high frequency, the interwire capacitance between the turns in the coil generally acts as a capacitive shunt across the entire coil. This shunt effectively shorts out the coil's output, hence the probe no longer functions properly.
The high frequency required in the above example dictates a reduction in the number of turns of the coil and an enlargement of the spacing between each turn. Other considerations require that the coil quality factor (Q) be greater than 1 and preferably be greater than 10. As is well known in the art, the coil quality factor is related to the ratio between the energy stored in the coil and the energy dissipated in the coil. The AC coil losses are functionally related to the energy dissipated and are dependent upon the diameter of the wire. For frequencies on the order of 7 megahertz, the wire diameter must be less than or equal to 2 mils. If the frequency is increased to 10 megahertz, a wire having a diameter of 1.4 mils is recommended.
The construction of a coil with wire less than or equal to 2 mils is very difficult due to the fineness and fragility of the wire. An additional factor to be considered is the geometry of the coil. In the carbon block example above, the precise location of the hair line surface crack is not known. Hence, the surface area to be covered by the coil is relatively large in relation to the depth of penetration of the eddy current in the material. As recognized in the art, the geometric configuration of the windings of the coil is important for the faithful reproduction of the coil. It is desirable to obtain at least two coils having similar levels of impedance. To meet the above requirements for the carbon block example, i.e., small depth penetration by the eddy currents, relatively large surface area, a relatively high excitation frequency, windings having a diameter on the order of 2 mils, and uniform but relatively large spacing between each winding (to minimize interwire capcitance), the coil must be substantially planar. A planar coil induces an eddy current in the carbon material over a large surface area and if the excitation frequency is high, at a relatively limited depth into the sub-surface region of the carbon block.
Experiments have shown that a fine wire, having a diameter of 2 mils, 44 gage AWG, placed on a planar insulating surface, cannot be replicated with sufficient accuracy. Thus the coils constructed by this method had impedances varying from coil to coil by approximately 10%. The variation in impedances between these experimental coils is not acceptable when the coils are utilized to detect flaws in materials. It is known that planar, unilaminar coils have been utilized as inductors on printed circuit boards. However, those printed circuit board inductors do not necessarily generate uniform electromagnetic fields proximate their coils.